The use of computers to simulate the functions of complex biological macromolecules is essential to achieve a microscopic description of biological processes and to model and interpret experimental data. Here we apply theoretical computational approaches to investigate the fidelity of T7 DNA polymerase, divided into discrete steps that include contributions from substrate binding, pK(a) shifts, and rate constants for the PO bond-breaking and bond-making processes. We begin by defining the discrimination between right and wrong nucleotides in terms of the free energy landscape for the dNMP incorporation reaction. We then use the linear response approximation and the empirical valence bond methods to obtain converging results for the contribution of the binding and chemical steps to the overall fidelity. These approaches are successful in reproducing general trends in the observed polymerase incorporation fidelity. The calculations demonstrate the potential for further integration of theoretical and experimental studies to analyze high- and low-fidelity DNA polymerases.